Skip to Main Content
Mobile Menu


PhD Dissertation: L. Damon Woods (U Idaho)

Investigators: R. Budwig (U Idaho), Kevin Van Den Wymelenberg (U Oregon), Elizabeth Cooper (U Idaho) and John Gardner (BSU)

Sponsor: Integrated Design Lab

Systems with significant delays, such as buildings with extensive thermal mass, are particularly problematic for most traditional control schemes.  Thermal overshoot and undershoot often occur, causing inefficient building operation and insufficient occupant comfort.  Model Predictive Control is a technique has been proven to be well suited for this class of systems, assuming an adequate model is available on which to base the control.  This project will use a combination of building energy modeling and on-site measurements to explore the impact of simplified building energy models on the efficacy of MPC for buildings with concrete floor hydronic systems.

Dynamic Modeling and Observer Design for Vapor Compression Cycle Systems

MS Thesis: Travis Pruitt (Boise State)

Advisor: John Gardner (Boise State)

Sponsor: DOE Industrial Assessment Center

Building on the previous project (Luthman & Gardner 2016) this project seeks to develop validated computer simulations of real-world HVAC systems.  Through a process of linearization and observer design, we hope to develop a methodology that will allow us to assess the instantaneous efficiency of the unit with a minimum of measurement.  For example, we could develop a real-time meter that will report current Coefficient of Performance (or EER) by measuring the temperatures of the evaporator and condenser coils and the current drawn by the compressor.


Schematic and Temperature-Entropy Diagram of typical Vapor Compression Cycle

Low Order Linear Building Energy Models for Commercial Buildings

MS Thesis: Sean Rosin (Boise State)

Researchers: Elizabeth Cooper (U Idaho), Jinchau Yuan (U Idaho) and John Gardner (Boise State)

Sponsor: Avista Corp

Using advanced parameter estimation methods, we are developing and validating low order linear models of building energy dynamics from both Energy Plus models and actual building performance data. The models are useful for load prediction, efficiency diagnostics and optimal control of building systems.

Agent Based Modeling for Residential Demand Response

MS Thesis: Ryan Schwartz (Boise State)

Advisor: John Gardner (Boise State)

Sponsor: Boise State/CEERI

Using a variety of simulation tools including Matlab/Simulink and AnyLogic, we are investigating semi-autonomous and self-organizing behaviors that will allow a loosely coupled network of residential thermostats coordinate useful demand response characteristics. This can create a more stable grid as well as accommodate a greater penetration of variable generators (i.e. wind and solar) on the grid.

Microgrid Design and Operations for a Campus Residence Hall

MS Thesis: Naland Johnson (Boise State)

Advisor: John Gardner (Boise State)

Sponsor: Boise State/CEERI

Boise State is committed to sustainability both in its educational program and the broad research agenda being carried out by our faculty. This project seeks to take our commitment to sustainability to the next level by engineering a micro-grid system that will allow us to make one of our residence halls completely carbon-free, not just carbon neutral. In other words, through a combination of renewable energy sources, energy storage and advanced control infrastructure, we are developing a plan which will allow students to experience the challenges and rewards of a 21st century, 100% renewable energy future.

D-Map view of actual electricity demand of a campus residence hall and the average daily profile

D-Map view of actual electricity demand of a campus residence hall and the average daily profile

Data Driven Modeling for the Advanced Metering Infrastructure

Investigator: John Gardner (BSU)

Sponsor: CEERI

The rapid expansion of the Advanced Metering Infrastructure provides unparalleled opportunities for energy management.  In addition, the proliferation of sub-metering technology in industrial applications and commercial buildings and campuses provides an even richer data set. Yet we have only scratched the surface of what is possible in this area.  I would characterize our current situation as drowning in data when what we need is information.  There are many opportunities to extract meaningful information from this ever-growing data set.

The figure to the left shows three plots of electrical energy consumption (in kWh) of a residence in Boise plotted against heating degree days (HDD) and cooling degree days (CDD). The bottom figure uses monthly utility bills but eliminates those months for which there were both significant heating and cooling requirements. Note that only a few months are available for which cooling dominates the load (shown in blue), greatly reducing the significance of the curve fit. Alternatively, the top graph shows the same data aggregated on a daily basis: more data, but muddier results. Finally, the middle plot, aggregated by week appears to tell a much clearer and internally consistent story.

This approach can form the basis of an easily-implemented algorithm to identify homes that would (or would not) benefit from additional weatherization.  It can also help create tools to better predict potential savings for improvement measures as well as verify the efficacy of previously applied measures.

Thermal modes of a multi-zone office building associated with two different time constants (Gardner, et al., 2013)

Thermal modes of a multi-zone office building associated with two different time constants (Gardner, et al., 2013)

The figure to the right shows an analysis using an approach similar to structural modal analysis where we track HVAC zone temperatures instead of nodal displacements.  This approach could not only help identify the magnitude of thermal mass in a building, but also the optimal approach to storing and retrieving thermal energy as part of an intelligent demand response program.

Steady-State Performance Models of Commercial Heat Pump Systems (Completed)

MS Thesis: H. Luthman (BSU)

Advisor: John Gardner (BSU)

Sponsor: CEERI

Refrigeration and air conditioning systems are ubiquitous in Western Culture. Therefore, it’s all the more surprising that there are few options for engineers when it comes to modeling the performance of these systems in off-design conditions. The complex environment of phase changes that occur in the evaporator and condenser coils creates significantly different heat transfer characteristics in different operating conditions. This project builds on some ground breaking work by Andrew Alleyne at the University of Illinois (Li and Alleyne, 2010) to develop both dynamic and steady state models of vapor compression cycle systems.  A significant contribution of this work is the ability to use standard performance data that is provided by the manufacturer to derive key model parameters through nonlinear optimization methods.

Our models are an important enabling technology that can be used in many relevant applications.  For example, they can form the basis of a system to monitor the coefficient of performance in real time with little or no additional sensor requirements above what is normally included in a chiller system.  Alternately, they can be incorporated into residential or commercial building system energy models to study the effects of demand response on various types of systems and home construction.  Finally these models can be part of a powerful and important education and outreach program for food processors and other industrial users who have large refrigeration loads, enabling them to maintain their systems in their most energy efficient states.

Targeted Energy Management Toolset for Comfort and Savings Based on Advanced Computational Intelligence Techniques (Completed)

Principal Investigator: Craig Rieger (INL)

Co-investigators: John Gardner (BSU), Milos Manic (UI), Kevin Van Den Wymelenberg (UI)

Industry Partner: Idaho Power Company

This work focuses on integrating advanced real-time visualization and control tools with HVACR systems, many of which employ direct digital controls (DDC).  The goal is to promote the achievement of substantial energy savings while accounting for the comfort and productivity of building occupants.

To develop a practical approach to implementation of technology, enabling better acceptance as well as consistent application, this effort will develop a user-friendly integrated toolset for use by the building managers in implementing, monitoring and maintaining their HVACR systems.

The primary objective of this work is the development of a toolset that integrates advanced sensor systems with advanced computational intelligence, e.g. neural networks and fuzzy arithmetic, to enable effective energy management while minimizing negative human comfort consequences and promoting improved occupant satisfaction.